Injecting some numbers into the AGI debate – Windows On Theory
[Yet another “philosophizing” post, but one with some actual numbers. –Boaz] Recently there have been many debates on “artificial general intelligence” (AGI) and whether or not we…
[Yet another “philosophizing” post, but one with some actual numbers. See also this follow up . –Boaz] Recently there have been many debates on “artificial general intelligence” (AGI) and whether or not we are close to achieving it by scaling up our current AI systems. In this post, I’d like to make this debate a bit more quantitative by trying to understand what “scaling” would entail. The calculations are very rough – think of a post-it that is stuck on the back of an envelope. But I hope that this can be at least a starting point for making these questions more concr
related reading
- Will scaling work?dwarkeshpatel.com
- AI 2027ai-2027.com
- Three Observations - Sam Altmanblog.samaltman.com
- Why AGI Will Not Happen - Tim Dettmerstimdettmers.com
- My picture of the present in AI — LessWronglesswrong.com
- I. From GPT-4 to AGI: Counting the OOMs - SITUATIONAL AWARENESSsituational-awareness.ai
- Scaling: The State of Play in AIoneusefulthing.org
- Artificial General Intelligence Is Already Herenoemamag.com
- The Scaling Hypothesis · Gwern.netgwern.net
- AI Timelines — LessWronglesswrong.com
- Clarifying and predicting AGI — LessWronglesswrong.com
- AGI is still 30 years away — Ege Erdil & Tamay Besirogludwarkesh.com